Major Chinese tech firms—including Alibaba, ByteDance, and Tencent—are continuing to pursue Nvidia’s AI chips, even though Beijing has strongly discouraged such purchases. These companies are keenly tracking their H20 orders and preparing for the potential release of the next-gen B30A model
Ongoing Demand Under Pressure
- Despite warnings from Chinese authorities, no outright ban has been issued. Companies are being asked to explain their orders, but they continue procurement.
- Nvidia’s H20 chip, specifically tailored for China under U.S. export controls, is priced between $10,000–$12,000, while the anticipated B30A is expected to cost double, yet potentially delivers up to six times the performance. Many firms consider this value compelling.
Why Chinese Companies Prefer Nvidia
- Domestic alternatives from Huawei or Cambricon remain limited in performance, prompting reliance on Nvidia’s unmatched AI capabilities.
- Nvidia’s own CEO, Jensen Huang, has reassured clients about availability, citing robust inventory and production readiness, including hundreds of thousands of H20 units and B30A samples slated for September.
Broader Industry Context
- Beijing is ramping up green energy and security-focused regulations, which may further complicate Nvidia chip usage. Financial Times
- The persistent demand for Nvidia chips underscores the tech rivalry between the U.S. and China and highlights the complexity of achieving semiconductor self-reliance.
Quick Summary Table
| Aspect | Details |
|---|---|
| Key Companies | Alibaba, ByteDance, Tencent |
| Chips in Demand | Nvidia H20 and potential B30A (Blackwell-based) |
| Government Stance | Strong discouragement, but no formal ban |
| Domestic Alternatives | Limited performance from Huawei and Cambricon |
| Market Outlook | Sustained demand signals reliance on Nvidia amid tech competition |
Conclusion
Despite mounting pressure from the Chinese government, top tech players are continuing to pursue Nvidia’s AI chips—emphasizing both performance advantages and limited domestic alternatives. The situation encapsulates the tug-of-war between technological independence and operational imperatives in the global AI landscape.
